From prediction to understanding: Will AI foundation models transform brain science?

IF 15 1区医学 Q1 NEUROSCIENCES

Neuron Pub Date : 2025-10-22 DOI:10.1016/j.neuron.2025.09.039

Thomas Serre, Ellie Pavlick

引用次数: 0

Abstract

Deep-learning approaches using massive data have transformed AI and are reshaping science. We ask when AI foundation models will transform neuroscience, outlining critical success conditions and a shift from prediction to explanation-linking computations to mechanisms of neural activity and cognition.

查看原文本刊更多论文

从预测到理解：人工智能基础模型会改变脑科学吗？

使用大量数据的深度学习方法已经改变了人工智能，并正在重塑科学。我们询问人工智能基础模型何时将改变神经科学，概述关键的成功条件，以及从预测到解释连接计算到神经活动和认知机制的转变。

本文章由计算机程序翻译，如有差异，请以英文原文为准。

求助全文

约1分钟内获得全文求助全文

来源期刊

Neuron 医学-神经科学

CiteScore

24.50

自引率

3.10%

发文量

382

审稿时长

1 months

期刊介绍： Established as a highly influential journal in neuroscience, Neuron is widely relied upon in the field. The editors adopt interdisciplinary strategies, integrating biophysical, cellular, developmental, and molecular approaches alongside a systems approach to sensory, motor, and higher-order cognitive functions. Serving as a premier intellectual forum, Neuron holds a prominent position in the entire neuroscience community.